JoAnn M. Holloway1, Mark P. Waldrop1, Kate M. Scow2, and Richard P. Dick3. (1) United States Geological Survey, (2) University of California, Davis, (3) Ohio State University
Among the barriers integrating microbial communities into ecosystem dynamics is our understanding of the extent to which compositional differences exist among microbial communities in different biomes, and how these differences translate into modifications of ecosystem function. This concept is addressed using mineral soils collected along N-S (Manitoba to Texas) and E-W transects (along the 38th parallel) through the U.S. Geological Survey and Geological Survey of Canada Geochemical Landscapes pilot study. Phospholipid fatty acids (PLFA; n=182) and enzyme assays (n=251) evaluated in conjunction with organic carbon and soil-water extract chemistry show distinct patterns of microbial biomass, composition, and function that are related to vegetation and land use variables. Gram-positive bacteria are more prominent in the central plains and Great Basin, with greater diversity associated with Gram-positive bacteria in these areas. Gram-negative bacteria are more prominent in the northern and eastern Great Plains, deciduous forest, and mountain forests in the Appalachians, Rocky Mountains, Sierra Nevada and California Coastal Ranges. This pattern is consistent with that of organic carbon, PLFA biomass and arylsulfatase. Arylsulfatase is low in regions with low annual precipitation, including the Great Basin, Colorado Plateau and southwestern Great Plains. Environmental factors, (e.g., temperature, precipitation, organic carbon, dissolved phosphorus) do not individually show strong correlations to microbial variables, reflecting the dependence of microbial community structure on multiple factors. Principle components analysis shows distinct PLFA microbial communities associated with different biomes. Generalized land-use patterns show increased microbial biomass and composition in agricultural soils relative to forest and grassland soils. Thus, continental-scale sampling reveals broad patterns reflecting the overprinting of land-use patterns over biome-dependent microbial community structures.